#Sony PlayStation
1 messages Β· Page 11 of 1
that was m4 x 16
well there was a fun creaking noise when screwing it together
haha
and some excess plastic i had to cut with a razor blade
well now you need M4 x 12 I guess because an M2 is probably gonna be a hotdog in a hallway
I have two other prints that I havent touched yet, I was going to try to get those to work
I'd try M2 x 12 or 11
the shaft is exactly 12 so if the screw is slightly long there's a chance it won't tighten... but if it's got a pointed tip it might dig in anyway π€·ββοΈ
Might try m2 x 12, having a hard to find 11
Let me know how it works out
lol, I could Dremel the end of the screw off
well yeah true π
I wanna know if 12 works though haha
I can't make it much longer or it infringes upon the memory card opening
but I got it to 12.25 in the new version lol
well if the 16 only gave me that much of a gap, then maybe i'll be okay
16 in theory would give you a 4 mm gap
might be about right lol
I can't even imagine where I got those numbers
I must have been on crack
(again)
would PLA have shattered in that instance? I ended up opting for PETG
hard to say lol
I've never tried to put a 4 mm screw in a 2 mm hole
that web site has stuff like nylon and is still pretty affordable
I thought about just snagging prints from there to sell but I don't think I'd squeeze out any profit
I think I might check that one out in the future. not super impressed with CraftCloud so far
Your analog pocket slot cover came out pretty good though
What would be the best material to print those?
I am now printing somehting in PLA Gold Colored to match my gold case
just about anything really, best would probably be ABS? Depends on your definition of best π
Gold PLA would be good
Should I be worried about the 0.8mm wall thickness warning on the top piece?
say what now
I'm guessing it's saying that about some of the lettering?
I don't even know where there would be a wall thinner than that
be warned the fancy top is a bear to print, supports get really wonky lol
unless you have great bridging then you might be fine
I tinkered with it and got it printing mostly decent
machine screws would work too? not just self tapping?
Cool, making a home depot order now
ouch
amazon sells them for cheaper π
even if you order 10 amazon is cheaper π
link?
Amazon is nuts, you can either get 20 sizes, or one size in 500 count lol
yup
but if you want a handful too bad lol
amazon is nuts and bolts π
well, number speaking, a handful. Like if you want 10 lol
Yeah, I don't need screws frequently enough to need the assortment. And I only need 10
Thanks for the help though
I'm full with bad ideas and useless information - follow me for more handy tips
subscribe
please use M3 screws for everything 
all Prusa owners have a bunch
is that PSX SNAC one made in Fusion 360 or some other free/cheap program? @tall rover
imo the design would be nicer if maybe the top was a separate piece that attached on and mine the screw goes through the corners. Printing that overhang seems like supports might be needed
Yeah Fusion
Not sure, it's @harsh shale 's design. I just had it printed from an online service. 3D printing is still a bit mysterious to me. I'm still trying to learn
So like a 3 piece sandwich? lol
indeed
This blue disc is where I had to end the screw shaft or, as you can see, it would hit the memory card area
yeah you could slice it yourself and glue it lol
the whole thing could be bigger or just use glue. Yeah slicing would work or separate lid that is gluable with alternate blank version would allow easy customization
You could make it bigger sure, I don't like to pad out the walls of something if I don't have to
You need supports to print it either way
The STEP files are available though if you want to do a remix!
yeah I think just change where you consider the top to be and you can slice in a way to not need supports
Eh?
it's got 2 holes stacked I don't know how you would avoid supports
without 3 levels and glue lol
Or padding the corners out
but lower
I canβt draw properly on here
like right at the bottom of the hole under the line
or the design is completely different and the panel with holes attaches to the side and the rest of the box is printed standing up tall
would look clean
That would make 3 pieces
you'd need glue
Panels to the sides? I suppose you could, that's a lot of work to avoid some supports
I gave in to just not have the headache, but also got the adapter to use the official snacs
Yes but then all future humans donβt need supports.
But then they have extra work printing, or an extra seam
If you want you can do a remix of the design in the way you're thinking and if it works out better that could be the new standard π
True but they'd need glue and would now have an extra seam, or they'd need to print 4 extra plates and glue those on and have further seams
To me it doesn't seem worth the trade off but I could be convinced otherwise
Iβm imagining two parts and no glue
I'd love to see how, without support
maybe I will try. Totally possible it wonβt work but sounds fun to try it
Sure!
And the STEP files are included
Windows 3d builder let's you slice STLs too if you just want to experiment
I really had to mull this exact problem over when I was designing it, so a better solution would certainly be welcome
For something ontopic: thanks to @manic citrus we now have single bin/cue in the unstable mister main build. I would celebrate it with a CD rip of my old audio CDs but i don't have a CD drive anymore, so a conversion from mp3 played with ridge racer must do it π
Thanks @manic citrus !!! π»
what does that even mean?
has bin and cue finally merged together to their final form?
I assume it means if you merge multiple bins into a single one then it will continue to work without issue.
A lot of old PS1 emulators wouldn't work with multiple bin/cue so I assume a few people had a catalog stored in that fashion.
you're an old emulator!
CHD all the way;
frameskipping across the pavement with your walking chair
Speaking of which, I haven't used a PS1 emulator since...2006? Those were the bad old days.
@tropic stratus Steel Harbinger freezes on fire, 100% repro, all settings default. I tested it with the dual ram nightly as well as release, all blessed bios. The second you press fire once you're in the game (the button works fine in the menus) the entire game hard locks. I've tried it on two MiSTers, a dual ram and a single ram analogue, both lock up.
I have only tested one chd (PSX Champ from IA).
OK, I guess it was the image. I found another iso and repacked the CHD. Worked fine. False alarm.
thanks for the fast response, really saved my day π
Hey I don't know if it's too early for this, but as far as I'm concerned you have a practically complete product. Congratulations, truly a job well done!
It's still showing more flakiness than any other game. To get it to behave I had to move it to the SD card. External SSD wasn't cutting it.
Ah, USB storage huh
it's not just a product of your imagination
Please Subscribe thank you!
Sony offered a preview of what gaming will be like in the year 2078 with the PlayStation 9.
Featuring electronic spores, retinal scanning, and a mind-control system, the PlayStation 9 is everything you could ever want from a game console. #PlayStation9 #1999
Yeah, the drives are super fast (film spec - I just happened to have them) but obviously usb2.0 speed cap.
Honestly I bought a 256GB low profile USB drive and put my Amiga and AO486 core folder on that, and have everything else on my SD
I wouldn't use a usb based storage for cd based games, as tempting as it is considering how much space they use
Isn't it possible to connect an external drive via CAT5?
I've never had any problems before. This game is very sensitive to speed issues I'm guessing.
Yes, but I haven't set up any CIF shares.
CFI?
Whatever the TLA is.
CIF
You'd be surprised. I use CD games over samba file share all day long without issues lol
The MiSTer handles the traffic very well
Again, it's this one game.
Zero other problems.
It's a Minscape FMV and honestly, it's probably the universe trying to protect me. The cheese factor is through the roof.
lol
But it has a somewhat famous actor in it and I wanted to get some footage of it. Edward Albert was a great b-movie actor and has a pretty interesting career.
Come to think of it.. original hardware probably couldn't do much better than usb2.0 speeds, so maybe that's not the issue
I'm not sure, but on SD no problems. I'll try to set up a share and see if that shows issues tomorrow.
I think SD has a higher minimum speed?
And the FMV itself is totally fine. It's only gameplay, which isn't streaming off disc.
But sound might be.
In fact, sound almost probably definitely is.
I looked at a few specs and they seem to both max out at 20MBps. USB2.0 seems to be the bottleneck on both storage solutions.
well, hopefully the quality of the image is completely ruled out as a factor
But microseconds are in play and response reliability may be an issue. These were PC folks working on new hardware. Mindscape that is.
It seems to be. The first 'always freeze' one works perfectly on SD.
I preserved it to test. But it's terrible on USB. The new one is much better, so I'm thinking the first one is still not a good rip.
Oh well, fixed for my purposes. I don't see myself playing this one for fun. Though it's not as bad as I would have expected. It's still a locked perspective 3rd person shooter with only digital controls. So it was never going to be great.
USB drives probably have more latency, which is a much more important factor due to the CD drive has to act in real time, otherwise the game might fail. There are measures to pause the core when the storage/linux cannot keep up with the speed, but this pausing has limits, e.g. when FMVs run, pausing is not possible
CHD decompression consumes another chunk of processing power, further increasing the latency
the sequential transfer rate on the other hand is more or less a don't care
The old emulators that supported CD drives often wouldn't work as well with USB CD drives versus IDE drives
Probably not as much a latency thing there, but maybe it was because this was before you had computers with enough ram to store the whole cd
I personally don't consider using real cd's that desirable anyways. They get scratched, ISO's don't.
that would really be the best solution: enough RAM to store the whole image. The transfer could happen while the bios shows and in worst case the core would pause if it's not fully finished
unfortunatly we don't have enough ram
Yeah, storing the whole image in ram is a setting you can turn in in many software emulators of CD based systems and it usually helps them run better
If we just had 1GB more :p
second sdcard could also solve this, but honestly i don't feel like adding(implement, test, debug, maintain) a second data path for the few who would want to use that
Yeah doesn't seem to be worth the effort.
especially since it works fine from normal SDcard or (wired) NAS
I even use it on wifi to a Nas and it works fine as long as the Nas always keeps the drives active.
Only occasional hiccups
But I manage my expectations with that.
You added safe pauses and stuff that makes it not crash the game I think.
yes i also have that with my NAS if a game doesn't use the CD for several minutes, then the core might pause for 10 seconds while the drives start to spin, but usually nothing bad happens
You can do a hack to keep the drives spinning, but the drives won't last as long.
Or just have a script read a sector just before the normal timeout period.
i added the safe pauses, but they are not 100% safe, as the core cannot pause any time. In worst case it can miss 1-2 sectors. Games that cannot stand that might fail. E.g. there was a football game that will freeze instantly if it ever fails to read one sector from the background music. Wouldn't want to play that with a real disc...must be 100% scratch free π
Thankfully I don't play football games.
Going a bit off topic here... I set my NAS to never spin down the HDDs. This is bad? Discs are in this state for about 5+ years.
ime I do the same thing to my home NAS and over the years have admin'd datacenter level storage arrays with the same deal, no spin down. I'd say it's more about power savings than anything, if a drive is marketed for NAS it should be built to sustain being spun 24/7. I'd worry if you were running more consumer drives though. In my previous home NAS I had wd red nas drives spinning for 5+ years and the dude who I pawned that one off on is still spinning the same drives some years later.
I also run raidz2 (sorta analogous to raid 6, I can sustain 2 drive failures before the 3rd one causes data loss) and have drives from different batches to hedge bets, so I'm not terribly worried if some drive fails after 5-7 years of being on
Its goooood https://youtu.be/bAojxWZRVKk
Lucky someone stopped him...
Shameful self plug! https://www.patreon.com/xcirrex
Green screen version: http://imgur.com/gallery/nJJZrMz
As someone that turns everything in my flat off when i leave for more than 4 hours except the Fridge... HOW CAN YOU SPIN!!! π
samba is MUCH faster than mister's usb2.0
fair enough yeah I forgot how poor the USB was
it's apparently faster than the sd card too
but i don't read chat history so i can't know that
good thing im here to set you right

Depends on the drives. If you are using enterprise drives or NAS drives then it should be no problem. If you are using deshelled external drives or things like WD Greens/blues then it could potentially be an issue because that method only half works
Usually those drives don't officially support staying spinning so this method would randomly spin them up at an interval over and over, multiple times a day. The unnecessary start-stop can wear out the mechanical components involved if they aren't rated for that many cycles like the enterprise/nas drive components are.
You still likely wouldn't see a problem for like 5-7 years anyways.
They are for NAS so should be safe then.
And I remembered that the NAS software have some bug that causes the hdds to spin up from time to time. So I find better to leave them always on.
Anyone finish silent hill on mister? After killing a late game boss and after the cutscene, I get a black screen. Background audio continues to play
this one?
if yes, this seems to be a rare bug in the game itself, see the linked thread. If no: please try to reproduce and create a savestate some seconds before the bug occurs
And yes: silent hill was finished several times now, according to reports in the old psx channel
Unless you're using a USB network device right? π
pretty annoying... happened twice in a row and i just gave up lol
USB2.0 is quite old, I'm surprised it's still common
That's interesting. I have beaten it several times on the PS2 (PAL version) and didn't encounter this.
on the ps2 can also happen
But it is definitely a bug in the game
I guess I must be one of the lucky ones?
yeah π
poor Cybil
Hi @wanton crown ! How is your 32sdram? π
still not received.
Sent with royal air mail with no tracking by misterfpga.co.uk
it's been sent for more than a week now, should have been there by now but nope
it's how the horror genre was designed back then
I was scared it was a psx core bug hehe
Afaik this was reported some time ago and it turned out to be a game bug
Was it after the fight with the police woman?
the old github bug more or less tells all about it
no idea how high the chance is to hit that bug
Some people have had problems with this on real hardware, you can read about it in several threads
yes, that's what i mean, not sure how high the chance is on real hardware. Maybe it's higher on the core/emulators? don't know, there is not much research about it. Just last time people tried to reproduce it, out of 3 testers, no one got the hang, so i guess for ~25% chance π
I hope we get some new core news soon. Feels like many people work on new things now. Not much mister news in January π€π
As things get more obscure or complex, development time increases. I might have some update this weekend, but nothing really exciting yet
Well, aside from bug fixes for obscure games, is there much else going on?
for psx? no. I only spend maybe 1/3 of my time in january on psx and i would be happy if it's even less next month, so i have more time for other things π
Agreed! The most exciting thing here is that the core is so solid right now, with features we never originally expected. π
now, what might be a cool feature, is if in the games directory, the core could treat cd images in a directory called "_bootcds" differently. so you don't have to open and close the lid if you want to boot a gameshark cd first, and then switch to another game, and it won't make a memory card entry for it
you ask too much!
LMAO
Dusts off Sonic 06
my ps brook adapter just stopped working so if i want to play my ps1 now i need to use my ds 1-3. #nothanks #mistersavedme
#notreallyplayingps1anyways
I still have my childhood psx, but I think the PSU is toast. light doesn't turn on and nothing works #mistersavedme
that sounds exactly what the noreset.txt is for
Yeah , noreset is the way
Yeah you're not hot
those days are long gone
any new bug reports ? , unexplained issues ? π
Hi all, and Hi @tropic stratus I want to thank you for everything you are doing with the psx core. I was shocked for what you were achieving and basically I have a Mister Fpga for your work. I see that you are reaching the end of the development and I want to ask you how is it going with the Captain Commando's issue. For what I know it's one of the few that doesn't have a workaround and harder to fix from what the creator of Xstation said.
yeah legendary Captain π
isn't that on cps2 anyway?
it is cps2
how come people want to play the psx version? bonus features or better music or something?
I don't think the PSX version is better than arcade
by the way , as far as I know , to fix this you need to implement a cd controller, similar to the one in xstation, and this will not happen
yeah i think rob said its basically a wontfix for now
there possibly isn't even enough room in the fpga to do the whole cd controller
yes , it requires Motorola MC68HC05 CPU
as for me it is not worth for two-three games to spend several months of work on it
plus retesting everything from scratch, since such a change would require it
but that's just me , I prefer to play the arcade version π
but im' sorry because it was not a question to me, I am not an expert
captain commando could be fixed on the core, it's more or less known what to do. The major drawback is: it breaks every savestate that has CD actions active(e.g. background music) and it would risk introducing several new bugs. That's why i haven't bothered yet to do it. Best is to implement the real CD controller some day in the future, which automatically solves this problem
breaking savestates is probably ok as long as you warn people
but yeah its a lot of faff for one game
to be honest: i'm currently not motivated to implement this additonal logic for just one broken enemy graphic in a single game and also have those drawbacks and spend like 1-2 weeks for it
we really have a better alternative to this game, an arcade version that works great
yeah i agree - also i don't want you distracted from your new project whatever it is π
our CD controller emulation is known to not be 100% perfect and while i fixed some bugs on it recently, i still think that having a true HW implementation is the real way to go to get this on the right track. The main problem is the effort/reward problem for it
required space is no issue, the 8bit CPU needs less space than the current emulation
huh i wouldn't of guessed that
well, we currently have dedicated logic for all the functions instead of some software. I would assume the size of the CD module to be half or less with the real CPU
the main problem with the CD controller is that the only software that runs on it is the CD controller firmware. How is one supposed to develop and test this CPU? it would require to set up a system to test the CPU , maybe even with handmade hardware for probing it...
also there is a lot of tight coupling with the mechacon and the XA decoder, all of which is not really documented/known to the degree that is needed
i hope this explains why i did the CD emulation approach. With the real CD controller, i don't think we would have many playable games yet on the core
(not to mention all other CD cores we currently have also do CD emulation, even further away in HPS)
Its CPS1
Hopefully sufficient documentation can be found, to hopefully make that step in the core's development easier. Looking at it from the angle of preservation. But yeah I too am more excited for future projects.
there is indeed documentation from no$ and it is by no means bad. Still it's not enough to build a cycle accurate cpu from it
the documentation is very valueable for all the connections from the HC05 cpu to the other parts
ooh a patreon update
@tropic stratus You keep us in suspense π
its a font rendering core
Me, trying to find out what console or handheld the Robert's new core will be in the Patreon update
if you print out the post, grind it up,roast it, and snort it on a full moon the new core will be revealed to you
well, at least you can be sure that it's a system that can render 2 colors minimum!
π
If thought it will be the virtual boy but I saw a video about it yesterday saying that there are maybe 3 good and interesting games on it. So I'm not sure anymore 
whatever it is i guess it has both homebrew and licensed games
most systems have at least some homebrew tho i expect
Just saw the patreon post of what I assume is a Sega Niomi core. Really excited π
Thank you π
"Hitachi SH-4 64 Bit RISC CPU (200 MHz 360 MIPS / 1,4 GFLOPS)" well.....no π€‘
it's sound processor alone could mimic a GBA easily π
π
Has feedback from the community been a factor to why you chose this new core? Or is it the core that you found the most interesting?
I like your approach Robert. First software emulator to understand the system
i must work on something where i burn myself for. I learned that with wonderswan, where i only ever really played 1 game. I was excited for GBA before because i knew it and have nostalgia for and for psx because it has 3D games
Do you have enough documentation for the system?
that's why i will likely never work on any arcade...i just don't like the type of games and i have no past with them
burning yourself I assume means you're excited yourself for it
maybe the term sounds wrong in english. Yes it was intended to mean: i'm excited to play games on this system
yes π
Great!
I just assume whatever it is can play DDR
N64-rumble-pack-core (snac required)
3DO core finally!!! π₯³
I am happy for any new core , no matter what system it will be β₯οΈ
Damnit, where there went my hopes for a Virtual Boy core 
but considering what is already implemented console wise... not thaaat many candidates I think.
Someday we'll get to it
It's a MiSTery!
Indeed 
Even fewer when you consider the ones Jotego has called βdibsβ on
but psx was also semi-dibbed
Yeah, in a way - it was already partially developed, but Laxer either a) ran out of steam, or b) never really intended to use the MiSTer framework and merge it into main
Jotego doesn't have "dibs" on anything..
Jotego just claimed NGPC as his
He's got so many irons in the fire that it's silly for him to say, "I'm going to do NGPC and Jaguar at an undetermined time in the future." It keeps other devs from considering working on them.
But that's just my offtopic two cents.
He's working on NGPC for over a year now. Today he said it's almost done. I'm not sure how the devs see this topic but I think it doesn't make sense to start a core when someone else claims he's almost done.
I think the dibs part was about Jaguar
I understood the reason jotego wants to do a jaguar cored more being that his highest paying Patreon supporter want him to do it.
NGPC was said to be done 18 months ago...sorry, but i don't understand that whole "one core" thing anyway. For emulators we have plenty for each system, why is that such a big topic for FPGA cores?
I don't know, honestly your cores are my favorites, so I would not mind as long as you're having fun doing it π
Youβre asking why NGPC matters?
At this point NGPC is the missing handheld - even Lynx would seem important if it was the only one missing
me neither, and it would be included in MiSTer proper too
I wonder if he'll release NGPC like 4 games a week
I wonder if heβll release NGPC
it doesn't really make sense to have multiple cores for the same system when the parts are interchangeable
all it does is split development for contributors and confuse users and annoy them with which to use
there are plenty of software emulators that could exchange parts even easier
then they should
it sucks trying to figure out what emulator to use for what
ever try to play NES on retroarch? like 12 different forks of fceux
it depends a lot on what you want to do. some just want to do the best emulation and are fine with improving the work of others. Some don't like the style that exists and want to write it different. Some want it to be "theirs"
none of which you should actually ever use
we're fortunate that we're in a position where we dont need to have such a situation
but what you propose are forks....if people start to improve things, others will improve different and we get different forks
ideally improvements are ported upstream
realistically what scenerio do you see in fpga where having two of the same system makes sense?
why do we need two lynx cores?
We have 2 PSX and 2 GBA π
lets take 1 example: i made a version of duckstation that is nearly cycle-accurate but much slower. It doesn't make sense to have that as general version, so it's useless. Should i delete it? I say no
much slower in a sense that modern systems don't reach 100% speed?
yes, 10% max
you can keep it for research but there's no reason it should be in the distribution if it's not as good
There's now two MSX cores.
for the time being
one of them will eventually go away
it's something we'll eventually fix
because it's not an actual MSX core
I hope the new one gets better and better.
It just has to reach feature parity to archive the old one
i don't see a problem for mister to only have 1 core per system, but i don't see any issue for several cores to exist overall. the "best" can be distributed via update for the normal user, but parallel development of another one can still be useful
the old one is de-facto deprecated
sure, that's fine
it's better to look at fpga as parts rather than systems though
a 6502 is quite mobile
mister is just mame in a box
it's more like mame's design philosophy, yes
if you want to make new versions of parts, that's fine
sure, there are premade components you can plug together, but even then there are shortcomings. E.g. the one in NES now has savestates, others do not. Even if you get a perfect core for something, it might miss things you want or maybe you want it in another language, ...
in a more tooling oriented world things like 6502 would be sub-gits
if a core is missing a feature, then add a feature
if a core has an innacurate part, replace the part]
if the core is just so rotten that most parts need replacing, or it's written in vhdl, then just rewrite it and replace the old one
in any case, the original comment was about starting the development of a core, while another one for that system is in private development. And i don't see any reason, why this should be a problem
It's not really
even if it would be open source, i would still consider that absolutly fair, but if it's closed, it's like non-existing for me
yeah, this is pretty true
yea, this idea of having a decent enough abstraction (that the verilog compiler can optimize away upon final βlinkβ), where you can have a common interface for data $, instruction $, and prefetch, so that you could have a package manager to plug in differing implementations (or keep a specific impl pinned of up to date)
that sounds like just getting high on your tools
most well made fpga parts are more or less pin accurate
swapping them out is usually not a big deal
well, lowering the barrier to entry for both new cores and existing cores
and making it easy to try different approaches
we duplicate the code to keep them self-contained but a lot of the commonly used parts should really come from a common github
watching the grafting of cache and memory interface onto PSX, and the work on the Amiga accelerator scene, makes me wonder if there can be easier cross-pollination but also easy to try things
yea, totally, could use sub modules
it's already easy
regardless the parts we use come from decades and many authors and places
expecting such a uniform set of interfaces isnt realistic
Tbh letβs say hypothetically Jotego and Robert both developed a Neo Geo Pocket and Jaguar core individually. Iβd still prefer Robertβs cores everytime bc of the save states and the extra mile QOL features that he adds
it's rare we need more than one version of a part anyway though
typically one is a clear winner
I would assume one might have a higher chance of being released as a 2P version and that would be nice for NGPC
with standard parts it also depends on the language. Mixing a verilog/vhdl project isn't great if you want to simulate it
so it depends on the development style. Some simulate, some don't
i can try to ask ChatGPT if it can check my verilog code and give me the warning Quartus cannot π
for glue logic, the conversion is easy. More typed VHDL code can be really difficult to convert to verilog
if you use vscode it can lint with icarus to give you a lot more strict warnings
does it accept typical verilog code we have in mister? I tried modelsim, but most emu level code throws so many errors, it's not useful
so no warning like "used before declared"?
generally the things that stuff like modelsim complain about, so you can write more compliantly
it does warn for that but you can configure which warnings you want
like for example it has some stupidly pendantic "all intendentation must be spaced" as a default warning
which I think is frankly wrong and disabled it instantly
Great, i agree with the tool π
tabs were literally designed for the sole purpose of indenting
so that people don't have to live with whatever mad coder uses 1 space indents
the problem with icarus will start when i map the port list to my vhdl entity
it likely cannot check for port sizes there
or even complains the entity does not exist?
I dont think it can do like intellisense cross language, no
T H I S!!!!
So glad you think this way and now we are all here on this channel! π
IMO if the core is old (like MSX) and it's a SOC instead of a proper original hardware, fine.
Otherwise I do believe it's an effort that could be put somewhere else. It's not like we have cycle accurate versions of every hardware on the planet already.
And as pointed out already by Kitrinx, improving =/= concurrent development.
IMO, if a person is motivated enough to make a core, just go ahead and do it.
anyone can make any core they want, but we dont need more than one available for mister. Whatever the best parts are that are are around should just get amalgamated into a single core
Seems like people are approaching the idea like it was software development.
It's Lego vs 3d printing.
Mister being the lego.
many (maybe even most) cores for mister have had one or more major chips in them replaced with better versions
nes, genesis, tg16, sms, among them
even jt cores are borrowing the 68k and things a lot, and likewise his sound chips are in a bunch of stuff
That's good to know. I had the impression jt cores where harder to mess with because the way jt programs them
they definitely are
but it's still hdl and the chips are still modular if you start from the pins
the salt is real. π€£
JT implementations actually were implemented in other places too, you know?
Also I believe saturn/CPS3 will both end up being better because of it.
God bless jt then
God bless EVERY developer that helps us build this community.
says user named "faithless"
Itβs a little bit funny that people are worried there could be more than one NGPC core under the circumstances
is there any known issues with memory cards lately? I have games that i know has some saves on it but it tells me to put in a formatted card in slot 1
No one has reported such problems, nor have I encountered anything like this
Virtual mem cards ?
Obligatory did you remember to open the menu after saving, before switching the MiSTer off?
(Iβm sure you did, but you gotta get the dumb questions out of the way)
well. it's first day i booted up the mister since the holidays so i dont really remembe
but i have saved games since before
im using that "autosave-thingy"
when you put your chd in a folder it creates a save with same name
Then try saving the save again in this game ?
And see if everything is ok
Delete the memory card assigned to this game
And what game is it?
it seems that i can't mount anything in slot 1
if i mount in slot 2 my Tomb raider save pops up
same file in slot 1, nothing happns
this is .sav-files
You donβt have SNAC enabled do you?
yes. is that bad? π
So you must use real mem cards
can you use real cards?
that's nice to now π
Snac and virtual memory cards cannot be used at the same time
what's wrong with you people? stop hooking up your freaking physical CLUMPS of memory cards and save onto the sd card like you're supposed to in 2023!
it's about to get physical in here!
as it is now, i cant save on sd-card since I use SNAC?
no idea, i don't use snac.
You can save on the other controller slot, if the game allows it
And youβve only got one SNAC controller enabled
Tomb Raider only allows on slot it, what i've noticed
Don't listen to him, he's jealous you have a snac and he doesn't π
have you heard of bluetooth?
Personally I'm waiting for a LLAPI compatible version of this core - you get native controller support like SNAC and virtual memory cards still should work π
the retrogaming community is always "personally waiting for"
No , everyone is always waiting for something, life is waiting

I too sometimes wait so long for shipments from the UK
the PSX->USB adapters are effectively just as fast as llapi and you dont need to wait on a special core
Thereβs no point on use LLAPI over USB
samba is very finicky for some reason for me
Then make it work on slot1 with SNAC turned on. Else this is what Iβm doing
it's something wrong with your NAS then
I find that sometimes windows 10 is finnicky with samba, but I have no issue with retroNAS on Debian
Samba on MiSTer can't transfer files larger then 1GB...
craps out with unknown error
tried with different computers and both wired and wifi
interesting...
yeah, I have to use a usb stick to transfer all large files
Iβm on Windows 11. Maybe itβs time to use Linux on gaming night
probably some settings that are set by default now and they shouldn't be for gamesharing.
https://techcommunity.microsoft.com/t5/storage-at-microsoft/smb-compression-behavior-amp-settings-changes/ba-p/3610763
Heya folks, Ned here again. As you know from a previous post, we released an end-to-end SMB compression capability with Windows 11 and Windows Server 2022. You can learn all about it atΒ SMB Compression | Microsoft Docs.Β Β There was an idiosyncrasy in the initial design that we changed with the rele...
I can confirm samba being finicky on windows 10 and it has nothing to do with the NAS or anything else
especially if you access it through a VPN
windows 10 IS the nas in that case
if you're talking about mister?
or do you mean outgoing samba?
Brazilian samba
well you can connect to mister via samba too, but in this case, I was talking about mister connecting to a samba server
mister's hosted samba is flakey
but the client is rock solid
not talking about mister, and I was talking about being a client under windows
I had to write some custom scripts to circumvent the issues when connecting to a samba server hosted being a vpn
there would be a chance to do virtual memcard with SNAC: if the memcard command is started, the virtual memcard could wait if SNAC does respond for the command. If not, it could take over. But it's unsure if the game will tolerate the timing, so it may lead to undetected memcard or even lost data. That's why i never touched it. Too much trouble for the win.
Isnβt possible to route the command to the virtual MC or SNAC based on a setting that says if the user wants to use the MC via SNAC or the virtual one?
First byte from CMD says if the PSX is talking to the controller (0x01) or the MC (0x81)
it's possible, but not easy, this is what I said before π
basically by the time the core knows if it's a controller or memory card read the device (real or virtual) is already selected. So it's hard to separate them. The only way would be delaying selection of either until you know which kind of transfer it is then recreate the things the core already would have did.```
maybe it could be an unsafe option
i'm down to test it, would be a big big qol improvement for me
maybe just live with the consequences of doing needlessly inconvenient things π
you want the real thing or not
it's not the latency actually its the negcon
no real way to plug it via usb and have it work right
I assume I can just use the PS3 memcard reader and convert that to a format that the PSX core can read, right?
er yeah i think someone made pc software to use that
you just need raw dumps, thats all the "format" is
Selection after first byte should be fine if the delay is not atrocious
if selection happens too close to actual data transmission things will fail and controller won't ack on that byte
Lower latency is a point. Anyway the blisster board works in normal USB mode on this core so it mostly works fine.
LLAPI is slower than fast polling USB in some cases
yeah its sort of been shown that it's not really all that fast is what i was getting at
you can get usb adapters in the 1.5-2ms range
'sort of' - where's the data on this then?
llapi is around .8ms
usb is .7-2ms
its not worth going out of your way to wait for a new core for <1ms
gasp. a whole .1ms slower. my money is wasted.
im not saying your money is wasted, I'm just saying its not worth waiting on the core for PSX specifically where it has known good adapters that are in the same realm
i'm actually concerned though on the NeGcon/jogcon side of things though. I just bought those and waiting on their delivery and now reading above suggests they won't work properly on this core without SNAC π’
I got my NeGcon working fully with the ostent USB adapter
it was a bit of a brain buster getting it mapped initially but it was fully functional
oh okay. neat. well maybe the blisster in usb mode will be the same π€
I been wondering about those adapters. A bit skeptical I haven't heard of them and they were really cheap.
Not saying cheap makes them bad but I'm am cautious about such things when I haven't heard of the name.
@still palm could it work if we just run the virtual devices always in background and if the memcard is accessed (0x81), instantly switch of the select lines for SNAC and continue with virtual memcard until the register interface takes away the select?
the real memcard would receive the 0x81 and nothing after, it just looses the select line after that. Shouldn't hurt the card or the logic?
Its amazing that Silent Hill still holds up even today. Looks and plays great on this core.
In some ways, the horror genre has only become worse over time in my humble opinion.
it went downhill as developers kept missing the point of low resources management, puzzles and psychology horror
Also the controls got better so the horror and anxiety of being unable to easily direct your actions in game got less π€£
The tank controls were the scariest thing about those games..lol
"Ooo Resident Evil..that's cool..It's got tank controls Ahh shit that's terrifying!"
By the end of re2 I was sort of attached to tank controls. I have a fondness for them now lol
They're just so... quaint lol
I liked tank control π€·π»ββοΈ
I did not and do not have a problem with this π
Those fixed camera angles really added to the experience with many of the early survival horror games.
Yeah I think the fixed camera angles are so cool and make the tank controls worth it
They're a little wonky but imo not bad at all once you get used to them
But the fixed angles are a lost art in modern horror games
We may see a resurgence one day, but at the moment it seems that many think that to leave the over-the-shoulder/FPS view is a regression.
I think as much as I hated it. the tank controls were the best for a fixed camera. It's in tomb raider where the camera wasn't fixed that it really got annoying.
What's really funny is play those tank controls games a few weeks then go back to a modern game without them. Gotta retrain your brain again.
Desert Strike (strike series) born and raised.
I think the restriction in controls, as already pointed out, is part of the experience. Fixed camera (so you don't know what's ahead of you), limited ways to dodge/evade, it's a mechanic. Like playing Soulsborne games on easy would change the experience drastically;
Ahh good Amiga times π
Many cool games
I love the strike series and those controls there didn't bother me..it's realistic for a helicopter to control that way.
What bios for North America is the best for you ?
- Don't remember if it has the cd "winamp/wmp" effects while playing audio cds, though. That was nice back then.
Models That Includes Soundscope: -SCPH 700X -SCPH 750X -SC0H 900X -SCPH 10X
Models That Does Not Includes Soundscope: -SCPH 100X -SCPH 3000* -SCPH 5000* -SCPH 550X
Those downloaded by update all are 7xxx
Thank ya
It was possible to disable snac by deselecting it when a MC transfer was detected.
So I added OSD option when SNAC is selected for either port a new option should show that you can toggle to use virtual or real memcards when snac is enabled.
I had to rework how select was done, and delay virtual transfers by a cycle.
the change could affect both real and virtual controller/mc transfers. so test both.
Make sure I didn't break anything.
So is that like proposed? Memory card will see the 0x81 command and shortly after it's turned off?
or did you implement something else? In any case: great work!
i can try the normal memcard behavior later and i will try to turn SNAC on while the memcard is being accessed π
Did i started all this memcard-talk just askin if there was any known bugs with memcards lateley? π
I dont want to start any trouble π
there have been many questions on this topic π
ok, good π
i've been hoping for this feature since i first soldered together a snac adapter π
but back then everything was still changing so fast it would of been crazy to have this on top probably
yeah when new option is used as soon as it detects 0x81 command it deselects the real MC and selects virtual MC then starts transfer on next cycle. there is a short time where real MC is selected but nothing is done in that time so should be safe.
Seems fine so far, couldn't find any issue. Very good work!
We need to find a clear way to tell the users what the option does. And the big question: what is default? virtual or real card?
i'd say virtual as default once its proven safe, most of the snac adapters sold out there don't have the memory card thingy
yeah
i think it's the better solution for most users, but existing users will have to switch 1 time then, as the "default" changes. I already received hate mail for such a change in the past when the core wasn't even released, so i'm cautious π
heh, best to delete those and pretend you never got it π
haters gonna hate π
nothing to worry about , on the internet everyone is "strong"
I chose real MCs as default. thinking it would work as it was before but now with an option to turn them off
i could switch it around though
decide now because it will be really annoying for people to save the setting and then that gets broken by switching behavior around later.
i don't use SNAC myself, so it's hard to tell for me, but i guess many of those that use a real memcard currently would switch to virtual one if it works
just having unlimited cards available without effort is worth it
I was considering a MemCardPro at some point, but this seems to make that unnecessary now.
I currently use MemCardPro but like Freddo said above, most adapters don't have a memcard slot so it would make more sense to have the virtual slot as default. Those of us who have gone through the trouble of finding a route to use physical cards with SNAC shouldn't have a hard time figuring out how to change a setting to continue doing that.
There isn't much point to use a real memory card slot aside from for the pocketstation and reading the saves at first to then write to a virtual memory card.
If the traction of getting people into getting a MiSTer setup is still growing strong, having virtual as the default will keep the questions from rolling in as those people start experimenting with addons like SNAC and whatnot
Yeah, I agree it's better default behavior.
could it try to communicate with the real memory card and switch if no memorycard is found?
if the real memcard was not found, the time for communication is already over
Yeah, it's a super low latency spi connection right?
the issue is that timeout is like 8 times the normal communication time. So you have to wait for so long, that it's already to late to generate the internal message
or at least for some games it migth be...not a solid solution like what blue1 has build now
and don't let us start to speak about half working memcards...this could mess up all your savegames. Nobody wants that risk really
Any expectation that it could cause any issues with some random games to disable the memory card port on that front? I would assume not. Maybe an auto function could be added later one day as an unstable option.
But if it's an unstable option that is VERY unstable it's not desirable.
I was wondering about this too, could an unstable option to speed up memory card reads be added easily or not? Like make it go as fast as it can? Hehe
i think some glitchy grpahics or even a game freeze might tolerated with unafe options, but loosing sa savegame completly is worst case
memcard is mostly slow because the BIOS code is horrible
Ah
I remember it being slower and faster at different points in the earlier stages of development of the core so was curious.
Just slightly though.
It would be awesome if the no$ bios could be expanded upon to solve all of these problems and be 1:1 compatible.
One day in the future.
that's the main problem...games depend on it. E.g. Diablo streams the music from CD via CPU to SPU while memcard access is done
Ahh
so if memcard would be faster, it drops sectors. I tried it, because saving in diablo is super slow
I'm almost entirely theory crafting here btw, no one get hyped up please π€£
no worries, i'm always interested in such kind of improvements. Even if it's not feasible, the challenge is often fun. But in this case i gave up early π
I think you've covered pretty much every improvement hack that would be feasible or matters that would work on the de10-nano
The core far exceeds what I could've imagined possible.
it's not a bad idea, but not as easy as the switch. I think this way is probably more reliable
Having to use real memory cards is what put me away of the snac adapter initially lol
right now I'm using a simple generic name for the option. What do you think the name should be?
this shows up under pad1,2 in OSD after snac is selected. SNAC MC,No,Yes
Any idea for a better more descriptive name? it needs to make it easy for people to understand what it does
If SNAC is enabled then real MC should be default. If SNAC is disabled, then virtual MC is default.
I would argue Virtual MC should be default in both cases since the Majority of Snac adaptors dont have the memory daughter board
I wouldn't have bought an adapter with a memory card board on it if I knew there was gonna be a solution for virtual memory cards with snac.
I thought maybe someone would eventually figure something out but not this soon.
Asked in the controllers channel, but I may as well ask here tooβ¦Does anyone have a PSX>USB adapter that works with the ASCII grip controller?
I want to use it in more than just the PSX core which is why Iβm wanting just a usb adapter
well, even some days ago it didn't looked like we get such a solution ever π
Yeah I ask a few days B4 I bought the adapter and it didn't sound good..lol
I did look at it before, it seemed tricky but wasn't so bad. could switch just in time to make it work
Well it's not the first something didn't seem feasible and someone found a solution. I've learned to never say never.
what if it appeared in the controllers spinners alongside regular snac
so it'd be like, digital, analog, dualshock... snac, snac mc ...
This one? I'm pretty sure the RetroPad32 works with it as I have this controller and I've tested it myself π
oh yeahh, I forgot about the retropad32!
Just not sure if I wanna spend that much yet just to get the one handed controller working π
have you tried it on your generic blue adapter?
I donβt have that one anymore sadly
but I did see that someone did a pull request for a PSX daemonbite on githubβ¦I can always just make one
It should work on any adapter as I didn't see anything special protocol-wise when dealing with it
there's retrozord aswell
its maybe the best diy psx->usb adapter
has support for lots of obscure things
yeah idk why but it just refuses to with this:
Did you test the controller?
gotcha
The retrozord looks cool, but I am curious about how well the daemonbite works as I already have the right arduino for that, Iβd just need to get myself a PSX controllers extension cable
Iβll probably just make a daemonbite and see if it does the trick
yeah go for it if you have parts on hand
Have you guys tried Parappa The Rapper?
I am not sure the game works right, timing wise: I mean, I get disqualified on the first stage, while I could easily beat at least three stages on the original PSX console.
Maybe youβre getting old? π
Yeah , i tried many times, It makes no difference to me whether real hardware or psx core, it always goes wrong π€£
yeah parappa's timing is pretty jank regardless of original console or emulation
Use snac , use the org controller and then tell me the differences
It all works the same for me π
Even the remaster has the same problem
Using a 3.3v arduino? Or 5v arduino with a level shifter?
ooh..uhhh I guess I didnβt know it mattered. Itβs the pro micro same as the other daemonbites use, so thatβs the 5v one right?
yeah, I have the 5v board
Check the project. But PSx devices uses 3.3v. Might work with 5v but can also damage it.
@bright yacht It is right here: https://github.com/MickGyver/DaemonBite-Retro-Controllers-USB/commit/8403735ac354ebe1bb68d5790e764544ad267e1e
That said, I just realized it doesnβt even support analog sticks, so maybe I will want to go with the retrozord after all
No analog stick is not good.
I recommend you to use a bidirectional level shifter if doing the above adapter or the retrozord.
One with 4 ports is enough
any resource that can point me to how thatβs installed? Iβm a complete noob at this π
I may be able to take it to someone else, but I kinda want to finally just start learning how to solder myself
Just to confirm, if my PSX ISOs are in 7zips that are inside a 7zip, I have to extract them all to use on MiSTer, right?
see the main deamonbite readme for getting started with flashing the software to the arduino itself: https://github.com/MickGyver/DaemonBite-Retro-Controllers-USB
Then you pretty much just need to follow the wiring diagram in the readme file in the linked commit above.
you should be good then
is there a particular BIOS that should be used?
but yeah must be uncompressed
so I can't put them back in a .zip like with other cores?
bios-getter from updateall will get you the 3 required bios's
yeah zips dont work on cd cores
oh, i wasn't aware of a bios getter script
if you use update-all already its probably quietly sorted out the bios files for you π or you might have to enable it
well, i did the updater and it got all the cores and such, but i wasn't able to launch a sega CD game until i got a bios and put it in the folder with the right filename
should there be a bios-getter script in there?
oh its disabled by default
run update-all and press i think its UP during the first 5 seconds and it'll take you to its menu so you can turn it on
whichever direction it is it says onscreen
k, and running update-all again won't hurt anything I presume? I haven't done any custom configuration for any cores, just put games in the respective folders and some bios files
should be fine
i don't think it even overwrites existing bios files or anything, maybe excepting ao486
is there a particular way i should create the folders in /games/PSX/ for it to work right? I found the Champion Collection which splits it between single disc and multi-disc games, if I make a folder called Multi-Disc and then have folders inside there for each game, that works?
you want a folder for each game, even single disc games
With Sega CD it always tells me when it has a update for bios files and to delete them if I want to redownload them.
folders are how mister works out what each "game" is so it can assign a unique virtual memory card each
Sure, I will have a folder for each, but I was thinking of putting multi-disc inside a folder that's explicitly multi-disc
you can but there's no real reason to
so /games/PSX/Multi-Disc/Chrono Cross/ for example
sure that'll work
I was just thinking so i would know which were multi disc, but I guess i'd find that out anyways when it told me to insert disc 2
plus single disc games will just launch when you click them
and multidisc ones will open the folder so you can pick a disc
well maybe I'll just skip the multi-disc folder then and just put them all in /games/PSX
pretty much any layout is fine aslong as its one folder per game
i assumed it works the same as on a megaSD or MODE
we used to have to organize stuff into /jpn /usa /pal or whatever so it could load the right bios but its done automatically now
i got pier solar MD+ working without issue
yeah for me i copied my collection over from my xstation sd card and it just worked
oh you can do .chd compression if you're really tight on space, its very effective about 50%
yeah i almost bought a PSX with an Xstation the other day but someone else snagged it
i'd rather just buy a bigger SD card than mess around with it, unless i found the champion collection in .chd
i can't imagine i could stuff this 1TB sata SSD in the box here also eh? heh
you could convert that existing collection pretty simply but yeah i mean, SD cards are pretty cheap
thats been my approach too
you can use usb storage if you want, an ssd is massive overkill but it'll do the job super reliably
yeah i mean, 1TB sandisk SATA SSD is $50
i bought three of them, one for the dreamcast, one for the saturn, and one for a console i didn't have figured out yet
the ones that are just slightly too big to fit in normal area? yeah probably
yeah theyre like, the tallest a 2.5" disk can be i think
except for no particular reason, the stuff insides tiny
Alright so made it here. Should I... Enable everything that's disabled?
Thanks! I know how to do the flashing and I know where each wire goes, itβs just the soldering really. Also, if I need that level shifter, I need to learn where that needs to go since itβs not part of the instructions 
well i enabled all the things that were disabled and it's running now
Looks like it's working
you only needed to enable Bios Database but turning it all ons harmless
I look at it another way. Showing the fact I paid for and put together the monstrosity might in some way have pushed blue over the edge to improve the technology. 
sure, but the other stuff looked like things i want too
whatever happened to the clock tower problem?
Did the ram arrive?
We are probably more anxious about it than the affected user.
and here I thought there was already a conclusion
I believe the ram delivery was delayed.
IMHO with snac if it's not 100% unscrewed with wires it shouldn't exist
trying to interpret and dice up the signal is just a huge rabbit hole full of endless broken edge cases and defeats the purpose of having it at all if you're not going to make it a real native interface
if you want to dice up and interpret signals, that's what usb adapters are for
I am 42 and very physically & mentally active, so no. In fact, I play other games better than ever. Never was I such a good player for 8 & 16bit games as I am now π
Then it's the game.. In fact, Bust a Groove works great on the PSX core, which is also a rythm game which requires no lag.
parappa is fortunate that it looks cool as hell, because its a terrible rhythm game
Oh, and I was trying my best with the game... tsssk, tsssk, bad, bad Parappa...
turn on some cheats and just blast through it its a trip
but yeah i wouldn't call it satisfying to learn or play
a few years ago when I bought the remater I thought "surely they have improved this game" unfortunately I soon found out that the lag was terrible still π
Or I have such delays in my head , I don't know.
it was probably even worse if anything
I'm doing pretty well in other games haha
technically the original isn't lagging it just has absolutely bizarre timing it wants you to hit that has nearly nothing to do with the visual feedback it gives you
yes
i did manage to complete it back in the day after like weeks of learning
but the trick is to practically ignore everything it shows you onscreen, its stupid
bust a groove is a much much better game mechanically
reaction time sadly isn't what it used to be. I remember that I used to do loops in the Atari version of Kung Fu Master, now I'm baffled at how I managed that. π€£
@still palm @tropic stratus just wanted to say thank you two for looking into SNAC + virtual mem card.
I was completely content relying on a physical memory card and that was just how I was going to handle saving but Iβm really happy that I can play my dorky ass Densha de Go games and have all my saves on the MiSTer hassle free.
glad you find it more convenient. I was kind of tired of people complaining about it
Poking my nose where it doesnβt belong but really, people were complaining? Iβm of the mindset beggars canβt be choosers. The fact that SNAC works alone is incredible.
it's hard to make everyone happy
I knew people would complain before I ever released anything, first version didn't even split up port1/port2. it was snac on both ports
Well, weβre all huge fans and supporters of what you do. Even if the extreme minority is a little cranky youβve made countless more people happy. Even if it isnβt perfect, youβre working hard to improve the MiSTer and thatβs the most important thing. Youβre awesome, canβt wait to see what you do next!
π€π₯π€
thanks Robby, now please remove your lips from my butt π
sorry @serene nebula no offense. was just joking around.
i had just sort of accepted it as a limitation and changed my workflow to accommodate. Glad it's likely going to have an extra option available
It didn't hurt that I stumbled into a thrift shop with a pile of $3 PS1 mem cards
i think it's a safe and convenient option. it's nice to be to be able to use real MCs or pocketstation, but useful since most people got snax so a multitap is needed to use MCs.
I read your comments about it before. and tried to make it the best I could. but I couldn't test everything with every game. So I need your help, and depending on testing and feedback I'll decide to add it.
yeah shoot me a PM with some permutations that you need tested if you come up with some
I have SNAX + Multi tap and a retrocastle snac now
ironically haven't been playing it as much... been in 64 land after getting an everdrive for xmas
mainly just need to make sure nothing happens to saves when using snac and virtual card. real cards should work like they did before.
It's good for other people, with different HW to try different games.
I didn't mean to single out Kajoq, I meant everyone. it helps with you guys testing.
especially @serene nebula he's awesome π
Thank you. Iβm just fortunate to be able to be part of this wonderful community.
that's actually really cool, I knew we'd see the 7800 core used for real hardware eventually
it's just so good
wow super necro, but thank you π
sorry, I was searching for something else and read that
Why not a copy memory card option? I have over 15 memory cards and would love to load then into mister.
you can with the bios menu
just select snac for one port and virtual memcard on the other
you may have to create virtual memcards before yourself by either adding them to the sdcard or saving once in the game you want to copy
Ow that is fantastic. Didn't know it was already possible. π
I remember being in that bios menu often making important choices of which save to keep because the memcards were so expensive back then π
Another +1 to virtual memory card being the default if using SNAC. Its all part of the convenience of using mister. I mean I have a real Wii with GC memory card slots and real memory cards to put in and I still use Nintendont's virtual memory card option. Because its convenient.
i just bought a snac adapter for psx for lightgun games, i can't wait :). Doing the same thing with a PC/emulator is still a nightmare i think. Thanks for @tropic stratus, @still palm and every one involved for making this possible and so easy.
A partial PSX family photo
we need to talk about that plant
Ohhh, are those arcade boards based on the PS1 architecture? I see what games the ones on the left/right are but whatβs the middle one?
Its a Taito G-Net, a Konami GV/Baby Phoenix/Crypt Killer
and Super Puzzle Bobble / Hyper Athlete / Crypt Killer
G-Net is PSX+ levels of power, GV is basically 1:1 PSX with a SCSI CD-Rom and Crypt Killer is 1:1 PSX on CPU/GPU side with more RAM
So fascinating. I would have assumed they would all have identical boards since itβs based on other hardware architecture.
no Sony let people do all sorts of shit with the PSX hardware
that Crypt Killer board is a full custom Konami design that uses SCSI Hard drive loading...and they designed that entire board from scratch...JUST FOR CRYPT KILLER
konami loved wasting money lol
Wild! I wonder how much different the arcade game is from the PS1 version.
Little extra lighting but otherwise near identical
it's almost like kojima was perfectly in line with how they did things, and it was konami that changed, not him
still no ram lol
I should have ordered from mister addons rather than misterfpga.co.uk
you have enough de10s, the risk of a next(ish) day delivery from an amazon seller would have been negligible
Anybody played that Master Doom PSX hack? Is it a standalone download, a rom hack, need to build it myself?
Creepy PSX doom with dual shock support would be nifty keen
Well, I can't get it to load on MiSTer. So I guess I'll try Vita.
OK, so I grabbed the wrong download and thought it was only available in PSP packaged pbp. But it's a bin/cue and it loads from chd perfectly, plays great, and apart from mapping R3 to fire, it's great out of the box!
Good framerate, excellent control.
Yeah, this is great. Like legitimately excellent.
Thanks @dawn wind!
Thanks for testing it! Iβll have to download and give it a try tomorrow night - Iβve been in a PSX Doom mood for some reason and saw this posted β¦ somewhere?
It's shockingly good. This might be the smoothest Doom on MiSTer...
Wonder how the 2X core handles it
I'm using one of the dual stick builds, so it might have the 2x. Let me take it back to stock. I did have all the juice on.
Feels very much the same on 'stock'.
itβs fine if you play only based on the music and the timing of the button images growing
no lag in the gameplay actually π
2X is a separate core actually
Yeah, I'm just not sure which dual stick build I picked up. I did it manually but it's probably the regular one. I'll grab the 2X one, but I don't think there's huge room for improvement. This is super optimized.
But I was running with Turbo on high, whatever that is, at first. I returned everything to stock (1x cd, no turbo), and the game still ran super smooth.
It probably internally caps at 30hz
I was just going to say that....I verified it is exactly that. Hehe. And 2x introduces ALL sorts of graphics glitches.
The graphics glitches cleared up when I rebooted, but I did get some slight edge alignment problems that I didn't see in the regular core. That being said, I was probably looking harder on the 2x core for stability issues. Regardless, great freaking game! The difficulty is dialed up from vanilla for sure.
asDssw
Hi cat
π actually a bt keyboard that was paired to the wrong computer..
lol from the keys, looks like you were trying to move in a game
@wanton crown whatβs the RAM ETA?
I just download my ram, it's the easiest way π€‘
Sounds like a USPS problem
no it comes from the UK
but tbh I've been playing the mister on a daily basis and still have encountered no issues other than the one reported for Clock Tower
have you tried clock tower again?
yea after the recent new updates, changed nothing
In the psx core, I can see the vertical crop and horizontal crop. When do I want to use vertical (224/270) and (216/256)? And what does horizontal crop do?
Fixed Hblank as well as Fixed Vblank can help delivering correct aspect rations and keeping the screen in sync with e.g. shaking animations. Both also offer crop options for games that depend on CRT viewports to hide artifacts at the edge of the image.
224/270 - NTSC , 216/256 - PAL
Thanks
So I should have that enabled if I have fixed v blank on?
I am exclusively playing ntsc games so 224/270 it is
If you haven't seen bugs in any game then you don't need to
For the game AirGrave it is useful.
Because it permits to show some part of the game not visible and also not visible on real hardware.
The top of screen in game you can have some sprites and no background layer.
With this option sets the top background layer is visble
if you have fixed vblank on you often get black borders even above and below the image. Use the crop options to get rid of it. Also some games depend on the viewable area being smaller than they set themselves, e.g. because crt couldn't show all. In these cases vcrop and hcrop can be useful
if you want a crt-like experience: use vblank and hblank on and set hcrop and vcrop on to 224
if you want as less black borders as possible and maximum viewable area, the default settings (hblank fixed, vblank unfixed, no crop) works for 95% of the games. For the other 5% you need to optimize
Found a real weird issue. Using a dualshock 2 controller through the blisster board on the PSX core, core version is this https://github.com/MiSTer-devel/PSX_MiSTer/blob/main/releases/PSX_20230114.rbf.
In Twisted Metal 3 with default settings, when I fire in game, the music stutters and game intermittently pauses.
If you turn on the OSD for CD you can see it is constantly pausing for CD access.
The game is dumped as multi bin - this one: http://redump.org/disc/2485/
I tried the same using a ps5/dualsense controller through bluetooth and its perfectly fine. So I guess since this doesn't happen when using a dualsense/ps5 controller through bluetooth its not really a PSX core issue?
Also tried the same using a brook wingman ps2 connected to the blisster board and the same dualsense controller and yeah it stutters using that too.
Sometimes the controller seems to get stuck and inputs repeat as well.
So seems like a blisster specific issue?
try turning rumble off via the ini and seeing what happens?
rumble=0 in your MiSTer.ini
it seems the HPS CPU usage is increased for some reason, so CD response doesn't come in anymore or even controller status is not transmitted properly anymore
Yes setting rumble=0 in the MiSTer.ini solves the issue. Except now I don't have rumble π¦
the blister usb stuff is a bit...odd. I wonder if it can't deal with the rumble data properly
have you ever updated the firmware on that blister?
Yes its running the newest firmware on the P1 port I am using for the controller
Since this seems to be a blisster specific issue I'll continue the discussion on their discord server.
Protocol-wise it shouldnβt interfere. Itβs probably a power related issue.
maybe. If the USB is repeated on/off due to rumble causing power loss, this will also make the HPS busy
i had the same with a broken cheap dance mat that would loose USB several times per second. DDR crashes from it
Thatβs what Iβve thought
at least one revision of the blister firmware had a pretty severe issue with whatever hid driver gets used in more modern linux. so it's possible not all bugs have been ironed out
Good call! Blisster uses V-USB which is in fact an USB 1.1 software (bit-bang) library. Issues under linux are not new with that library and protocol could really be an issue here aside of power: https://forums.obdev.at/viewtopic0fd4.html?t=8865
I have been using v-usb for ages with Linux and MiSTer, no problems.
"no problems" is always "within limitations". USB v1.1 itself is pretty darn slow
Daemonbite uses same bitbanging too AFAIK?
Pretty much what you have to do with those tiny atmegas and Arduinos unless you add an extra chip I think.
I thought daemonbite was using an extra chip, or had USB built in
I donβt think so, but maybe someone who knows more can chime in?
Maybe you are right: https://docs.arduino.cc/hardware/leonardo
Adafruit was crowing about their super-low-end chips being able to do "bit banging USB", but changed their tune several years ago saying it was less and less compliant, so you might not want to do that anymore
From their "Trinket" page: Deprecation Warning: The Trinket bit-bang USB technique it uses doesn't work as well as it did in 2014, many modern computers won't work well. So while we still carry the Trinket so that people can maintain some older projects, we no longer recommend it. Please check out the Trinket M0. It has built-in USB, more capabilities, and is comparable in price!
It doesnβt make sense for any project to implement bitbanged USB nowadays
What is the difference between bitbanged USB and proper USB?
Ohh is bitbanging essentially protocol software emulation?
software emulation of the USB. Regular microcontrollers aren't fast enough to get the timing right. It's also a protocol which theoretically requires + / - values on each of the lines (though they are differential).
RP2040 can do bitbanged USB with PIO, but I still wouldn't advise it
A lot of microcontrollers in the past 5 years just have it baked in
(at least the signalling level; higher-level software protocols are usually layered on top)
To autoload an mcd it needs to have the same name as the folder/rom, correct?
Or are only .sav files autoloaded?
only sav
Can mcd be converted to sav? I know a person who transferred an mcd to the mister and is wondering how to autoload it
and it's only the name of the last folder + ".sav"
yes, just rename them
mcd has no header or additional data, just raw 128kbyte
Ahh so manually make a copy of the mcd and rename it to folder name + .sav?
yes. If the game is in games/US/abc/game.cue, then the save name will be abc.sav
Fantastic, thank you very much!
Question:
I have a large collection of eBoots conversions I use with various consoles. Any slight possibility you could implement eBoot loading into the core?
it's not bound to the core, but HPS side, so it's code which is not costing FPGA ressources but i doubt this will happen. The need for it seems very low and with the format supporting even encryption and compression it can be a lot of effort for one to write
Thank you!
I was also wondering if there is a new version of the PSXCPU2X or was that a one-off release? I use the update PSX script to stay on top of stable and unstable versions of the core.
you will not obtain this core by udpate all
Thanks Kuba! Think thatβs the one I have. If I recall correctly, Robert indicated it was just a test core and was not going to be actively updated? Wanted to make sure I hadnβt missed anything on that front and I just canβt get enough of this excellent developers work
there will be new versions of cpu2x core
should appear with the release of the official core
Thank you!
but yes it;s experiemntal build , so no official release π
Iβm having issues w/ swapping discs on Aconcagua. I like to test disc swapping on the multi-disc games and every gameβs working except this one atm. Iβm testing by loading the second disc first and then loading disc 1 when prompted. This game gives me the error βIncorrect disk inserted. Please insert Disk 1β
Open > Swap to Disc 2 > Close worked for me awhile back
i suppose that game's more sensitive to whether or not you swap your discs properly for whatever reason
Wow that worked brilliantly. Thank you!
no prob!
i uploaded a new testbuild with blue1 last feature, allowing to use virtual memory card together with SNAC on the same port. Would be great if it can be checked for issues, otherwise i would build a release from it the next days
Now I'm sorta annoyed that I got a snac adapter with the memory card slot coz I thought I would have to use it to save. I could have got 1 without the memory card slot that matched all my other adapters.
you can desolder the daughter board if you want
I thought it was only friction fit
depends on who made it
its still a neat thing to have esp if you're going to maintain a real PSX aswell
yep only issue is its not cased unless you use @dinerto design and have a 3d printer
It's not the memory card slot itself. It's that I wanted a enclosure for the board. I gotta get that separate. I could have got an adapter without the MC slot that came with a enclosure.
right
check if you can pop off the mc part and then get a case 3d printed somewhere?
I just launched discord after weeks and I see this. Iβm definitely going to try this Iβm so happy this is being tried as I play a lot of Time Crisis with SNAC. thank you so much to you and blue1. You guys keep on giving and giving youβre so amazing π€©
ah this is great! I've been disabling and re-enabling SNAC to use memory cards with my dualshock up until now. i'll give it a try later
I've noticed some weird behavior using a USB controller (DS4) with the "dualshock" option in the core lately also. it seems to default to analog being off, and if I enable analog in the BIOS before starting a game, it reverts to analog being off again, and loses player assignment for the controller
even thought it says "slot 1 analog" once i use the combination to toggle analog in a game, pressing any buttons on the controller do not register until I chose the "reset player assignment" option in the OSD
Figured I better explain the new change but hopefully it's self explanatory.
Whenever you enable SNAC on either port a new option is shown
"SNAC MemCard,Virtual,Real"
It will default to using virtual cards(for which ever port snac is enabled), but you can still use real ones by toggling the new option.
You can still backup or move saves between virtual and real MCs in bios if you only enable snac for one port.
This should give the best of both worlds, and give people the most options to play how they want to play with snac.
analog mode is often reset by games when booting them up, so this is no big surprise. Changing the mode is more like a soft request to the game, but the game can force any mode any time. What i don't see is what the player assignment has to do with it. The player assignment is completly in the HPS and has nothing to do with the core, so i can't imagine how the analog switch would influence it
It got messed up with those steps. I noticed while testing the SNAX Virtual Memory card build today it was turning off analog also. Tried the same steps there and it unsurprisingly didnt mess up player 1 input. Using USB it did though
I had no other controllers plugged in when that happened, only the DS4 and a keyboard that was far away
so for some reason the HPS looses the DS4? Can you still use the OSD with it?
yeah osd still worked, it just wasnt assigned player 1 anymore when the game started, for some reason
i had to use the "reset player assignment" function in the OSD to get any response from the controller
what happens if you force the controller type to digital? does it work then?
i generally use digital when using the core and have never noticed any issues with that
sorry, i mean when the DS4 is already in the broken situation
Tested the SNAC Virtual Memory card build extensively today, and everything seems to be working well on the memory card side
there was a graphical glitch in chrono cross i had never encountered though. i recorded a video
ive only had that happen once and would have to try and re-create the situation. i'll do so while trying to reproduce this chrono cross graphics glitch in the last official core build/release
video will be finished uploading in a couple minutes
i really wish i would have created a save state when i saw the glitch. i never think to use those lol
i can't remember how long ago was the last change to anything graphics related...maybe something was loaded wrong. We had a single glitch in FF8 when the dress of quistis was loaded wrong that was in the savestate but never reproducible
haha fun bugs to try and track down
could be original game behavior also with some of those super-random ones
but you use standard settings? i'm getting glitches too in FF9 when i use turbo high to overcome the slowdowns in battles
yeah standard settings. i dont use anything hacky generally
fixed vblank and hblank were on, and the fade to black option
ok, all fine than, was just asking. The square RPGs are tempting to use turbo and FastCD as they run so much better with it....unless you hit the bugs π
I'm always scouting for bugs when playing games on MiSTer. leaving on hack options and stuff like that makes it much more difficult to know if its the stock core behavior causing issues
i do love quality of life stuff like that though. especially great for CD rpg's with tons of loading like the FF4/Chrono Trigger ps1 game
